Fetal Hyperglycemia and a High-Fat Diet Contribute to Aberrant Glucose Tolerance and Hematopoiesis in Adult Rats

nature publishing group Articles Basic Science Investigation

Fetal hyperglycemia and a high-fat diet contribute to aberrant glucose tolerance and hematopoiesis in adult rats

Emily K. Blue1,2, Kimberly Ballman1,2, Frances Boyle1,2, Eunjin Oh1,2, Tatsuyoshi Kono1,2, Sara K. Quinney3, Debbie C. Thurmond1,2,4,5, Carmella Evans-Molina1,2,4–6 and Laura S. Haneline1,2,4,7,8

Background: Children exposed to gestational diabetes with GDM are more susceptible to obesity, type 2 diabetes, mellitus (GDM) during pregnancy are at increased risk of obe- hypertension, and the metabolic syndrome (5–8). However, sity, diabetes, and hypertension. Our goal was to identify meta- the cellular and molecular mechanisms that underlie the link bolic and hematopoietic alterations after intrauterine exposure between prenatal exposure to diabetes and disease later in life to maternal hyperglycemia that may contribute to the patho- are not well understood. genesis of chronic morbidities. Since glucose freely crosses the placenta to the fetal Methods: Streptozotocin treatment induced maternal circulation, manipulation of maternal glycemia is a strategy hyperglycemia during the last third of gestation in rat dams. commonly used to model diabetes in pregnancy in animals. Offspring of control mothers (OCM) and diabetic mothers The majority of previous studies focused on pregestational (ODM) were evaluated for weight, glucose tolerance, insulin diabetic models. These studies demonstrate that intrauterine tolerance, and hematopoiesis defects. The effects of aging exposure to hyperglycemia leads to altered glucose tolerance were examined in normal and high-fat diet (HFD)-fed young (9,10), bone development (11,12), renal function (11,13), and (8-wk-old) and aged (11-mo-old) OCM and ODM rats. vascular function (9,14). However, few studies have been con- Results: Young adult ODM males on a normal diet, but ducted that replicate the pathologic effects of GDM with onset not females, displayed improved glucose tolerance due to of hyperglycemia during late pregnancy (15–19). increased insulin levels. Aged ODM males and females gained There are a variety of environmental factors such as diet, more weight than OCM on a HFD and had worse glucose tol- exercise, and aging that may influence the development of obe- erance. Aged ODM males fed a HFD were also neutrophilic. sity, type 2 diabetes, and hypertension. In addition to intrauter- Increases in bone marrow cellularity and myeloid progenitors ine exposure to GDM, consumption of a high-fat diet (HFD) preceded neutrophilia in ODM males fed a HFD. is a risk factor for these chronic morbidities. Moreover, these Conclusion: When combined with other risk factors like diseases are common in older populations, leading to the sup- HFD and aging, changes in glucose metabolism and hema- position that aging also promotes disease pathogenesis (20,21). topoiesis may contribute to the increased risk of obesity, type Furthermore, many inflammatory cell types, including macro- 2 diabetes, and hypertension observed in children of GDM phages and neutrophils, contribute to obesity-linked systemic mothers. inflammation (22–24). We hypothesize that the combination of prenatal exposure to hyperglycemia, postnatal consumption of a HFD, and aging synergistically increase the risk of obe- estational diabetes mellitus (GDM) most often occurs sity and diabetes, partly due to the contributions of changes in Gduring the third trimester of pregnancy when maternal β-cell function and in hematopoiesis. To test this hypothesis, insulin resistance increases and insulin secretion is inadequate, we used an established rat model of GDM in which hypergly- leading to hyperglycemia (1). Between 5–10% of pregnancies cemia occurs during the last third of gestation. Previous stud- in the United States are affected by GDM (2), which doubled ies using this model demonstrated sex differences in metabolic from 1994 to 2000 (1,3). GDM is linked to many health com- and vascular phenotypes (16–18). Therefore, we prospectively plications in both the mother and fetus, including preeclamp- examined males and females for all phenotypes tested. Young sia, fetal macrosomia, and fetal death (4). In addition to and aged offspring from these pregnancies were challenged these acute complications, GDM exposure in utero increases with a HFD. In addition to evaluating for effects on weight the child’s risk of disease later in life. Children of mothers gain and glucose tolerance, studies were conducted to examine

1Department of Pediatrics, Indiana University School of Medicine, Indianapolis, Indiana; 2Herman B Wells Center for Pediatric Research, Indiana University School of Medicine, Indianapolis, Indiana; 3Department of Obstetrics and Gynecology, Indiana University School of Medicine, Indianapolis, Indiana; 4Department of Cellular & Integrative Physiology, Indiana University School of Medicine, Indianapolis, Indiana; 5Department of Biochemistry and Molecular Biology, Indiana University School of Medicine, Indianapolis, Indiana; 6Department of Medicine, Indiana University School of Medicine, Indianapolis, Indiana; 7Department of Microbiology & Immunology, Indiana University School of Medicine, Indianapolis, Indiana; 8Indiana University Simon Cancer Center, Indiana University School of Medicine, Indianapolis, Indiana. Correspondence: Laura S. Haneline ([email protected]) Received 20 December 2013; accepted 11 August 2014; advance online publication 17 December 2014. doi:10.1038/pr.2014.185

316 Pediatric Research Volume 77 | Number 2 | february 2015 Copyright © 2015 International Pediatric Research Foundation, Inc. Effects of GDM on offspring Articles whether alterations in circulating inflammatory cell levels and Maternal Hyperglycemia Causes Hypoglycemia in Male ODM hematopoiesis existed. Exposure to maternal hyperglycemia did not alter the birth weights of surviving males (OCM 7.1 ± 0.7 g, n = 63; ODM RESULTS 6.9 ± 0.9 g, n = 54) or females (OCM 6.7 ± 0.7 g, n = 64; ODM Streptozotocin (STZ) induces necrotic death of rat pancre- 6.6 ± 0.8 g, n = 54). To evaluate for neonatal hypoglycemia atic β cells resulting in insulin deficiency and subsequent and hyperinsulinemia, blood glucose and plasma insulin lev- hyperglycemia (25). To induce hyperglycemia in late gesta- els were measured in OCM and ODM on postnatal day 1 in tion to model human GDM, pregnant rat dams were injected random-fed neonates. Male ODM had significantly decreased with either citrate buffer (control mothers) or STZ (GDM blood glucose levels compared to male OCM (75 ± 10 mg/dl, mothers, 45 mg/kg STZ) on day 12 of gestation, similar to n = 9 vs. 87 ± 15 mg/dl, n = 8; P = 0.016). Plasma insulin lev- previous studies (16–19). Injection of STZ induced a rapid els in males did not reach statistical significance, but trended increase in maternal blood glucose levels by gestational day higher in male ODM (male OCM 0.6 ± 0.2 ng/ml, n = 8; ODM 15 (Figure 1a). To confirm a loss of functionalβ cells in 0.9 ± 0.4 ng/ml, n = 10; P = 0.09). Blood glucose levels were also GDM dams, pancreata were evaluated for insulin express- lower in female ODM compared to OCM (69 ± 9 mg/dl, n = 12 ing β cells. As expected, STZ-treated dams had reduced vs. 75 ± 6 mg/dl, n = 9; P = 0.049). Insulin levels in female ODM β cell area (Figure 1b, control dams and Figure 1c, GDM were not statistically different from female OCM (0.8 ± 0.7 ng/ dams). To assess for potential direct effects of STZ on fetal ml, n = 9 vs. 0.5 ± 0.3 ng/ml, n = 9; P = 0.33). β cells, quantitation of pancreatic β cell area was performed. In contrast to data in the mothers, no differences were Young Male ODM, but Not Female ODM, Have Increased Glucose observed in β cell mass between offspring of control mothers Tolerance (OCM) and offspring of diabetic mothers (ODM) neonates In humans, children of mothers with GDM have an increased (Figure 1d,e). Quantitative analysis of the insulin-posi- risk of obesity, insulin resistance, and the metabolic syndrome tive β cell areas revealed similar β cell islet areas for OCM at young ages (5–8). To determine if hyperglycemia during and ODM (3.4 ± 2.0% for OCM and 3.0 ± 2.1% for ODM; late gestation in rats led to any of these metabolic changes, P = 0.82). Dam weight gain was lower in GDM mothers weekly weight checks and monthly glucose tolerance tests compared to controls (Table 1). There were no differences were conducted in young (3–13 wk old) OCM and ODM. No in litter size, but an increase in maternal deaths during labor differences in weight were detected in OCM and ODM rats was observed in the GDM mothers (P = 0.009, Table 1). On during this time. Interestingly, a small but significant increase postnatal day 1, pups from citrate- and STZ-injected dams in glucose tolerance was seen in young male ODM compared were crossfostered to mothers who had not been subjected to male OCM (Figure 2a,c). In contrast, there were no differ- to STZ or citrate injection during pregnancy. The OCM and ences in glucose tolerance between OCM and ODM females ODM were evaluated for weight gain, glucose tolerance, (Figure 2b,c). Similar results for glucose tolerance were insulin tolerance, and hematopoiesis. observed at 4 and 13 wk of age (data not shown).

bc a 600

400

200 de Blood glucose (mg/dl)

0 12 15 18 21 24 Day of gestation

Figure 1. Streptozotocin (STZ) induced elevated glucose levels during the last third of gestation, due to β cell death. (a) After either injection of citrate (control, circles) or STZ (gestational diabetes mellitus (GDM), squares) at gestational day 12, morning blood glucose levels were tracked in the timed pregnant mothers. Graph shows mean and SEM glucose levels. (b–e) Immunohistochemical measurement of pancreatic β cells by insulin staining reveals loss of β cells in dams, but not for neonatal pups. Representative images are shown for (b) control dams, (c) GDM dams, (d) offspring of control mothers neonates, and (e) offspring of diabetic mothers neonates. Line denotes scale for 200 µm.

Enhanced glucose tolerance can be attributed to either data suggest that the improved glucose tolerance observed in increased insulin secretion or improved insulin sensitivity the young male ODM is likely due to an aberrant increase in of peripheral tissues. To examine for evidence of increased insulin secretion following glucose administration. insulin secretion, insulin levels were measured following glucose injection during a glucose tolerance test. These studies Consumption of a HFD Diminishes Differences in Glucose showed that ODM males have increased plasma insulin lev- Tolerance in Young Male ODM els in response to glucose injection compared to OCM males Unexpectedly, young male ODM displayed increased glucose (Figure 2d). Consistent with the findings for glucose toler- tolerance (Figure 2a). These rats were fed a normal diet with low ance, insulin levels in female OCM and ODM were similar fat content (14% kcal from fat). Since increased caloric intake (Figure 2e). To evaluate males for changes in insulin sensitivity, is linked to weight gain, obesity, and risk of type 2 diabetes, we insulin tolerance tests were performed. OCM and ODM males hypothesized that the dietary stress of a HFD would lead to had similar insulin sensitivity (Figure 2f). Collectively, these increased weight and impaired glucose tolerance, especially in the ODM rats. To test this hypothesis, a cohort of OCM and Table 1. Maternal data for rat GDM model ODM were weaned from their mothers to either the normal diet or a HFD (60% kcal from fat). Rats were weighed weekly to track Control GDM mother mother weight changes (Figure 3a,b). Consumption of a HFD over the 23-wk feeding period resulted in a significant weight gain in Total number per group 20 30 P value males (519 ± 34 vs. 441 ± 31 g, P < 0.0001) and females (258 ± 32 g Weight: GD 12 (g) 257 ± 22 262 ± 16 0.57 vs. 231 ± 19 g, P = 0.004), revealing significant effects of HFD on Weight: GD 21 (g) 346 ± 34 329 ± 31 0.21 weight gain (P < 0.0001 by two-way ANOVA). However, no dif- Maternal weight gain (g) 91 ± 15 68 ± 21* 0.03 ferences were detected between OCM and ODM. Litter size (live pups) 10.7 ± 2.1 10.7 ± 4.1 0.99 Glucose tolerance tests were performed on 8-wk-old rats to determine if ODM rats on a HFD had impaired glucose Maternal deaths during labor, n (%) 1 (5%) 5 (17%)** 0.009 tolerance. Similar to data illustrated in Figure 2a, the ODM Means ± SD are shown. GDM, gestational diabetes mellitus. males fed a normal diet showed slightly improved glucose tol- *P < 0.05 by unpaired t-test. **P < 0.01 by χ2 test. erance compared to OCM controls (Figure 3c). Interestingly,

ab600 600 c * 500 500 30,000

400 400 20,000 300 300 * 200 200 * GTT AUC 10,000 Blood glucose (mg/dl) 100 Blood glucose (mg/dl) 100

0 0 0 0306090 120 0306090 120 Male Female Minutes after glucose injection Minutes after glucose injection de f 2.5 2.5 150

2.0 2.0 100 1.5 1.5 * 1.0 1.0 50 Insulin (ng/ml) Insulin (ng/ml) 0.5 0.5 Blood glucose (mg/dl) 0.0 0.0 0 0246 0246 0306090 120 Minutes after glucose injection Minutes after glucose injection Minutes after insulin injection

Figure 2. Eight-week-old male offspring of diabetic mothers (ODM) have improved glucose tolerance due to increased insulin response to hyperglycemia. For all graphs, offspring of control mothers (OCM) is shown in white, and ODM is in black. (a,b) A GTT was administered to fasted 8-wk-old (a) male and (b) female rats. *P < 0.05 by repeated measures ANOVA followed by Sidak’s multiple comparisons. (c) Area under the curve measurements for GTT are shown. Significant effects of ODM were observed by two-way ANOVA. *P < 0.05 by Tukey’s multiple comparisons. OCM: open bars; ODM: closed bars. For panels (a–c) n = 8 males and n = 7 females. (d,e) Insulin levels were quantified in a separate cohort of OCM and ODM within 5min after glucose injection. Graphs show plasma insulin levels before and after glucose injection for (d) males (OCM, n = 8; ODM, n = 8) and (e) females (OCM, n = 7; ODM, n = 9). *P < 0.05 by repeated measures ANOVA followed by Sidak’s multiple comparisons. (f) Insulin tolerance tests were performed on 8-wk-old male OCM and ODM. Graph shows glucose levels at various times after injection of insulin (n = 4 OCM and ODM). No significant differences in insulin tolerance were detected using repeated measures ANOVA. Graphs show mean and SEM values. GTT, glucose tolerance test.

a 600 b 600

500 500

400 400

300 300 Weight (g) Weight (g)

200 200

100 100

0 0 51015 20 25 51015 20 25 Age (weeks) Age (weeks) c d 500 500 * * 400 400

300 300 * 200 200

100 100 Blood glucose (mg/dl) Blood glucose (mg/dl)

0 0 0306090 120 0306090 120 Minutes after glucose injection Minutes after glucose injection

Figure 3. High-fat diet (HFD) induces weight gain in offspring of control mothers (OCM) and offspring of diabetic mothers (ODM) males and diminishes differences in glucose tolerance in young males. After weaning, young males were fed either a normal diet or a HFD. For a-d( ), OCM is shown in white symbols, ODM shown in black symbols, normal diet shown in squares, and HFD shown in triangles. (a,b) Weights are shown for OCM and ODM (a) males and (b) females. Both OCM and ODM males and females fed a HFD show increased body weight compared to those fed a normal diet starting at age 10 wk for males and 11 wk for females (P < 0.05 by repeated measures two-way ANOVA followed by Tukey’s multiple comparisons). However, there was no significant difference in weight at these ages between OCM and ODM on either diet for either males or females. For males, OCM normal, n = 9; ODM normal, n = 9; OCM HFD, n = 8; ODM HFD, n = 8. For females, OCM normal, n = 9; ODM normal, n = 9; OCM HFD, n = 9; ODM HFD, n = 10. (c,d) GTT results for fasted 8-wk-old OCM and ODM (c) males and (d) females. Males fed a HFD show impaired glucose tolerance (*P < 0.05 for HFD effect by repeated measures two-way ANOVA with Tukey’s multiple comparisons). Females showed no effect of HFD on glucose tolerance. There was no significant difference in glucose tolerance between 8-wk-old OCM and ODM on a HFD. For males, OCM normal, n = 5; ODM normal, n = 5; OCM HFD, n = 5; ODM HFD, n = 4. For females, n = 5 for all groups. Graphs show mean and SEM values. GTT, glucose tolerance test.

Table 2. Weight gain over 2 mo in 11-mo-old rats Males Females Normal HFD Normal HFD OCM, n = 5 ODM, n = 5 OCM, n = 5 ODM, n = 5 OCM, n = 6 ODM, n = 5 OCM, n = 6 ODM, n = 6 Weight gain (g) 5 ± 11 19 ± 6* 79 ± 8** 118 ± 13*,** 2 ± 11 2 ± 5 21 ± 14 38 ± 27† Means ± SD are shown. Statistical significance defined as P < 0.05 by two-way ANOVA followed by Sidak’s multiple comparisons. HFD, high-fat diet; OCM, offspring of control mothers; ODM, offspring of diabetic mothers. *P < 0.05 compared to OCM males on same diet. **P < 0.05 compared to males on normal diet. †P < 0.05 compared to ODM females on normal diet. consumption of a HFD led to reduced glucose tolerance in tolerance in young females (Figure 3d). Thus, HFD consump- both OCM and ODM males, eliminating the improvement tion eliminated improvements in glucose tolerance of young in glucose tolerance observed in the ODM males on a nor- ODM males. However, HFD induced similar glucose intoler- mal diet (Figure 3c). In contrast, HFD did not impact glucose ance in young OCM and ODM males.

Aged ODM Have Increased Susceptibility to Metabolic fed either a normal diet or a HFD for 2 mo. This approach Derangements of HFD was chosen so that adult rats were fed a HFD for a similar We hypothesized that postnatal exposure to the combina- length of time as studies in younger rats and to avoid poten- tion of environmental stresses seen clinically (i.e., aging and tial developmental consequences of HFD consumption from HFD) would lead to profound impairments in the glucose the time of weaning. Consumption of a HFD for 2 mo by tolerance of offspring exposed to intrauterine hyperglyce- older males led to increased weight gain in both OCM and mia. To test this hypothesis, 9-mo-old OCM and ODM were ODM males (Table 2). However, the male ODM HFD group

a Males b Females * 500 † 500 *

400 400 * * † † † 300 300 †

200 200 Blood glucose (mg/dl) Blood glucose (mg/dl)

100 100

0 0 0306090 120 0306090 120 Minutes after glucose injection Minutes after glucose injection

cd* 50,000 * 50,000 * * 40,000 * 40,000 30,000 30,000

20,000 20,000 GTT AUC GTT AUC

10,000 10,000

0 0 Normal HFD Normal HFD

ef* 6 6 *

4 4 *

2 2 Insulin (ng/ml) Insulin (ng/ml)

0 0 Normal HFD Normal HFD

Figure 4. High-fat diet (HFD) increases weight gain in offspring of diabetic mothers (ODM) males and females and impairs glucose tolerance. Nine- month-old offspring of control mothers (OCM) and ODM males and females were fed either normal diet or HFD for 2 mo until the age of 11 mo. OCM: open bars; ODM: closed bars. (a,b) GTT results from (a) male and (b) female 11-mo-old OCM and ODM rats. OCM are shown in white symbols, ODM shown in black symbols, normal diet shown in squares, and HFD in triangles. †P < 0.05 for significant effect of ODM andP * < 0.05 for significant effect of HFD by repeated measures ANOVA with Tukey’s multiple comparisons. (c,d) Area under the curve measurements also showed significant effects of both ODM and HFD for the males and ODM and the interaction between ODM and HFD for the females. For (a,c) males, OCM normal n = 4, ODM normal n = 5, OCM HFD n = 6, ODM HFD n = 5. For (b,d) females, OCM normal n = 6, ODM normal n = 5, OCM HFD n = 5, ODM HFD n = 6. (e,f) Random-fed insulin measurements are shown for 11-mo-old (e) males and (f) females. *P < 0.05 by two-way ANOVA with Tukey’s multiple comparisons. For (e) males, OCM normal n = 3, ODM normal n = 5, OCM HFD n = 5, ODM HFD n = 5. For (f) females, OCM normal n = 5, ODM normal n = 5, OCM HFD n = 6, ODM HFD n = 6. Graphs show mean and SEM values. GTT, glucose tolerance test.

ab6-mo males 11-mo males c 11-mo females * 8 8 8 *

l) 6 l) 6 l) 6 µ µ µ / / /

4 4 4

Neutrophils (K 2 Neutrophils (K 2 Neutrophils (K 2

0 0 0 NormalHFD NormalHFD NormalHFD

de* f * 250 * 100 100 ) ) )

3 3 * 3 10 10 10 × × × ( ( 200 ( 80 80

150 60 * 60 * 100 40 40

50 20 20 BM progenitors/femur BM progenitors/femur 0 BM progenitors/femur 0 0 Normal HFD NormalHFD NormalHFD

ghi 50 * 50 150 *

* ) ) ) 6 6 6 40 40 *

10 * 10 10 × × × 100 30 30

20 20 50 10 10 Total BM cells ( Total BM cells ( Total BM cells (

0 0 0 Normal HFD NormalHFD NormalHFD

Figure 5. Male offspring of diabetic mothers (ODM) fed a high-fat diet (HFD) have neutrophilia and increased bone marrow (BM) progenitor cells. Neutrophil counts, bone marrow myeloid progenitors, and total bone marrow cellularity measurements were performed on the same animals at each age. Offspring of control mothers (OCM): open bars; ODM: closed bars. (a–c) Neutrophils (K/µl) were quantified by complete blood cell count a( ) in males at 6 mo, (b) in males at 11 mo, and (c) in females at 11 mo of age. Eleven-month-old ODM males fed a HFD exhibited neutrophilia (b), while all other experimental groups had normal neutrophil levels. For (a) males at 6 mo, n = 8 for all groups. For (b) males at 11 mo, OCM normal n = 3, ODM normal n = 5, OCM HFD n = 5, ODM HFD n = 5. For (c) females at 11 mo, OCM normal n = 5, ODM normal n = 5, OCM HFD n = 6, ODM HFD n = 6. (d–f) Myeloid progenitors were quantified per femur in d( ) 6-mo-old males, (e) 11-mo-old males, and (f) 11-mo-old females. All HFD-fed ODM males, and 11-mo-old HFD-fed ODM females had increased myeloid progenitors. For (d) males at 6 mo, OCM normal n = 8, ODM normal n = 8, OCM HFD n = 7, ODM HFD n = 7. For (e) males at 11 mo, n = 5 for all four groups. For (f) females at 11 mo, OCM normal n = 5, ODM normal n = 5, OCM HFD n = 5, ODM HFD n = 4. (g–i) Total bone marrow cellularity was measured in (g) 6-mo-old males, (h) 11-mo-old males, and (i) 11-mo-old females. Bone marrow cellularity was increased in HFD-fed ODM males and females. *P < 0.05 by two-way ANOVA with Tukey’s multiple comparisons. For (g) males at 6 mo, OCM normal n = 8, ODM normal n = 8, OCM HFD n = 6, ODM HFD n = 6. For (h) males and (i) females at 11 mo, n = 5 for each group. Graphs show mean and SEM values. gained significantly more weight compared to the OCM fed group (Figure 4b,d). Thus, in both males and females, the a HFD. For female rats, only the ODM on a HFD for 2 mo experimental group with the worst glucose tolerance was the exhibited significant weight gain Table( 2). Thus, in both 11-mo-old, HFD-fed ODM. Insulin measurements from male males and females, the combination of intrauterine exposure and female rats showed elevated insulin in ODM HFD-fed to hyperglycemia, HFD consumption, and aging induced the males and females (Figure 4e,f). These data indicate that pre- greatest weight gains. natal and postnatal factors contribute to the development of We next examined the effect of the postnatal stress on glu- insulin resistance and glucose intolerance. cose tolerance of aged OCM and ODM rats. Even on a normal diet, 11-mo-old ODM males exhibited impaired glucose toler- HFD-Fed ODM Males Exhibit Neutrophilia and Increased Bone ance when compared with OCM males (Figure 4a,c). While Marrow Myeloid Progenitors HFD consumption impaired glucose tolerance in both OCM To determine if there were differences in circulating leukocytes and ODM males, ODM HFD-fed males displayed the worst in OCM and ODM rats fed a HFD, complete blood cell counts glucose tolerance. For the females, impaired glucose toler- were performed. No differences in circulating leukocytes were ance was only detected in the 11-mo-old aged ODM HFD-fed detected in 6-mo-old rats (Figure 5a). However, an increase

Table 3. Adipokine levels in 11-mo-old rats Males Females Normal HFD Normal HFD OCM, n = 3 ODM, n = 4 OCM, n = 5 ODM, n = 5 OCM, n = 4 ODM, n = 4 OCM, n = 6 ODM, n = 6 Plasma adiponectin (ng/ 21 ± 6 18 ± 2 26 ± 5 28 ± 5 30 ± 7 28 ± 7 24 ± 5 32 ± 11 ml) Plasma leptin (ng/ml) 7 ± 2 6 ± 1 67 ± 34 189 ± 121* 3 ± 1 4 ± 2 6 ± 3 11 ± 4** Means ± SD are shown. Statistical significance defined as P < 0.05 by two-way ANOVA followed by Sidak’s multiple comparisons. HFD, high-fat diet; OCM, offspring of control mothers; ODM, offspring of diabetic mothers. *P < 0.05 compared to ODM males on a normal diet. **P < 0.05 compared to ODM females on a normal diet. in circulating neutrophils was observed in 11-mo-old ODM study, we show that the early improvements in glucose toler- males fed a HFD (Figure 5b). No difference in neutrophils was ance were associated with an inappropriate increase in insulin observed in females at any age evaluated (data for 11-mo-old secretion compared to controls. This apparent discrepancy in females are shown in Figure 5c). Neutrophils are short-lived findings is likely due to the timing of insulin measurement leukocytes that are replenished continually by myeloid pro- after glucose challenge. Our data demonstrate increased insulin genitor cells that reside in the bone marrow (26). To determine secretion at very early timepoints (5 min) while no differences if ODM males fed a HFD have altered myeloid progenitor cells in insulin levels were observed at later times (30 min) (18). in the bone marrow, the cells from the femurs of OCM and Moreover, our findings are consistent with emerging clinical ODM males were collected and quantified using standard clo- data that suggest that young children of mothers with GDM nogenic progenitor assays. At 6 mo of age, myeloid progenitors exhibit improved glucose tolerance due to increased insu- and total bone marrow cells were significantly increased in the lin secretion (32,33). Additional subtle differences that exist ODM males fed a HFD (Figure 5d,g). In 11-mo-old males, between our data and others (18) may be due to variability in consumption of a HFD increased myeloid progenitor cells maternal hyperglycemia as tight glycemic control is challenging and bone marrow cellularity for both OCM and ODM groups in the rat GDM model. Furthermore, a number of hypergly- compared to rats fed a normal diet (Figure 5e,h). However, the cemic dams in our study died during labor, which could have most dramatic increases in myeloid progenitors and total bone masked potential birth weight alterations detected previously. marrow cells were detected in the aged ODM males fed a HFD Another important point is the distinct processes for selection (Figure 5e,h). In females, alterations in myeloid progenitors of offspring for crossfostering, which included only the larg- and total bone marrow cells were only observed in 11-mo-old est pups in the prior report and was random in our study. In ODM females fed a HFD (Figure 5f,i). general, however, the findings between studies are comparable. Previous reports demonstrate that fat-derived hormones like Together these observations, in humans and rodents, support adiponectin and leptin regulate the growth of hematopoietic early β cell dysfunction in offspring of diabetic mothers. While progenitors and stem cells (27–31). We assessed adiponectin in utero, a hyperreponsive fetal β cell would be beneficial to and leptin levels in the serum of 11-mo-old rats. Adiponectin combat a hyperglycemic environment. However, continued levels were not significantly different between OCM and ODM insulin hypersecretion by the β cell would be predicted to have for males or females (Table 3). More striking were the changes detrimental effects, promoting increased adiposity and insulin in leptin concentrations. Eleven-month-old, HFD-fed ODM resistance (34–36). In addition, prolonged hypersecretion of males had significantly increased leptin levels, consistent with insulin may contribute to endoplasmic reticulum stress, result- the increased body weights of these rats (Table 3). In addition, ing in β cell exhaustion and/or apoptosis (37,38). 11-mo-old, HFD-fed ODM females had a modest increase To test whether external postnatal stresses potentiate the risk in leptin concentrations compared to controls (Table 3). for ODM to develop insulin resistance, rats were fed a HFD Collectively, these data link intrauterine hyperglycemia, aging, and allowed to age. While HFD reduced glucose tolerance and HFD consumption to increases in leptin concentrations slightly in young (2-mo-old) OCM and ODM males, a syner- and altered hematopoiesis. gistic effect of being a male ODM, aging, and consumption of a HFD yielded the highest plasma insulin levels and the most DISCUSSION severe impairments in glucose tolerance. Increased insulin Our data indicate that young males exposed to hyperglycemia levels in the context of impaired glucose tolerance infer that during the last trimester of gestation have altered insulin secre- the HFD-fed, aged ODM males were insulin resistant. While tion in response to glycemic challenge. Neonatal male ODM increased adiposity from consumption of a HFD is a well- were hypoglycemic and trended toward hyperinsulinemia, in appreciated contributor to insulin resistance (23), quantitative a manner similar to human neonates exposed to GDM (4). measurements of adiposity were not conducted in the current Glucose tolerance tests in young ODM males displayed a coun- study; so, direct correlations with detected metabolic pertur- ter-intuitive increase in glucose tolerance, similar to a prior bations are not possible and would be interesting to pursue in report in rodents (18). However in contrast to the previous future studies.

Normal diet: 3 wk–6 mo

STZ or citrate Birth injection Maternal HFD: 3 wk–6 mo GD12 hyperglycemia Crossfostering GD 15–21 P1–3 wk

Normal diet: 3 wk–11 mo

Normal diet: 3 wk–9 mo HFD: 9 mo–11 mo

Figure 6. Schematic of the rat gestational diabetes mellitus model and experimental procedures. HFD, high-fat diet; STZ, streptozotocin.

Presumably, early β cell dysfunction together with increased an important role for neutrophils in the initiation and prop- insulin resistance in male ODM enhances the lifetime risk to agation of obesity-associated inflammation (22,43,44). develop type 2 diabetes. Given these observations, it is interest- Neutrophils are mobilized from the bone marrow to initiate ing to speculate that the mechanism responsible for the early, an inflammatory response in an acute setting or for a sus- sustained alterations in insulin secretion may be due to aber- tained inflammatory response in chronic diseases (45). This rant epigenetic regulation of key molecules involved in β cell is achieved by neutrophils secreting cytokines, proteases, and function and/or development. Previous studies in an intra- chemokines that recruit other immune cells, especially macro- uterine growth restriction rat model demonstrated that intra- phages. Adipose-associated macrophages are well appreciated uterine growth restriction offspring had impaired epigenetic as contributors to obesity-related inflammation (24,46). The regulation of a key transcription factor, pancreatic and duo- expansion of bone marrow myeloid progenitors and cellular- denal homeobox 1, that regulates pancreatic development and ity may have been facilitated by high leptin concentrations. maintains β cell function (39). Furthermore, alterations in the Leptin is known to regulate hematopoiesis, and leptin recep- regulation of pancreatic and duodenal homeobox 1 expression tor-positive cells are important components of the bone mar- were linked to a decreased insulin secretory response to glu- row microenvironment (29–31). Future studies that examine cose and the development of type 2 diabetes. Whether a simi- the interplay between alterations in adipokines, metabolism, lar mechanism occurs in ODM is unknown, however our data and inflammation will promote our current understanding suggesting early β cell dysfunction in rat ODM prior to insulin of the predisposition of ODM to develop components of the resistance support the future interrogation of this question. metabolic syndrome. A limitation of our studies involves the possibility that STZ Interestingly, ODM rat females were relatively protected may cross the placenta and directly affect the fetal pancreas. from the changes in glucose tolerance, insulin secretion, and Although this rat model of GDM has been used in several pre- hematopoiesis observed in the ODM males. Our data are vious studies (16–19), direct assessment of this point has not consistent with previous studies in Sprague-Dawley rats that been reported. STZ, a glucose analog, is imported by β cells showed that females are resistant to diet-induced obesity (47). via the GLUT2 transporter (40) followed by induction of β cell Some modest changes in weight and glucose tolerance were necrosis (25). Experimental rationale for use of STZ to model documented in the ODM females, especially with aging and late gestational hyperglycemia without directly affecting the HFD. While increased bone marrow cellularity and myeloid fetus includes the short half life of STZ (40) and reduced fetal progenitors were observed in aged, HFD-fed females, neu- levels of GLUT2 compared to adults (41,42). While a possibil- trophilia was not detected up to 11 mo of age. We speculate ity exists that STZ crosses the placenta, we detected no evi- that further aging, additional stressors, or other genetic back- dence of β cell loss or reduced insulin production in neonatal grounds may be required for significant perturbations in ODM pups. metabolism and hematopoiesis to occur in female ODM. Significant alterations in hematopoiesis were detected in In sum, these studies demonstrate metabolic and hematopoi- ODM males on a HFD as early as 6 mo of age (i.e., increased etic alterations in offspring exposed to late gestation hypergly- bone marrow cellularity and myeloid progenitors). These cemia combined with aging and postnatal HFD consumption. abnormalities intensified with age and preceded the develop- Future studies to elucidate the molecular mechanisms involved ment of neutrophilia in HFD-fed, aged ODM males. A simi- in the evolution of these pathologic phenotypes are warranted. lar, albeit delayed, pathological progression was observed in Given limited animal data on the effects of GDM on offspring, females. Neutrophilia is extremely relevant in the context of our data enhance current knowledge and provide the founda- obesity-associated inflammation. Several studies establish tion for those future studies.

METHODS Plasma Insulin and Adipokine Measurements In Vivo GDM Model Trunk blood was collected from neonatal rats for glucose and insu- All animal studies followed the guidelines for care of animals at Indiana lin measurements. Frozen plasma samples were thawed and assayed University and were approved by Indiana University Institutional for insulin (from glucose tolerance test in 8-wk-old and from Animal Care and Use Committee. A schematic of the GDM model and random-fed 11-mo-old rats measured using a radioimmunoassay postnatal exposures is shown in Figure 6. Timed pregnant Sprague- kit), leptin (from random-fed 11-mo-old rats using a radioimmu- Dawley rats were obtained at gestational day 9 (Harlan, Indianapolis, noassay kit) and for adiponectin (from random-fed 11-mo-old rats IN). On day 12 of gestation, while under isoflurane anesthesia, animals using an enzyme-linked immunosorbent assay kit). All assay kits were injected in the tail vein with either citrate buffer (control moth- were obtained from EMD Millipore, Billerica, CA. Assays were per- ers) or 45 mg/kg streptozotocin (GDM mothers) (Sigma, St Louis, MO). formed in duplicate, and standard curves were performed with each Glucose levels were checked every morning from gestational day 15 assay. The limit of detection for the insulin radioimmunoassay was until parturition using an Accu-Check Aviva glucose monitor (Roche, 0.01 ng/ml for a 100 µl sample. Several baseline (time 0) values for Indianapolis, IN). An average morning blood glucose level of 8.9 mmol/l insulin levels in the glucose tolerance test were below the limit of (160 mg/dl) was set as the minimum for inclusion in the study. Ten per- detection. In order to complete repeated measures ANOVA, values cent of STZ-injected dams (3/30) had glucose values that ranged from of limit of detection/2 were used to replace the missing values for 4.5–8 mmol/l (81–148 mg/dl) and were excluded from the study since repeated measures ANOVA. they were not sufficiently hyperglycemic (1). Ninety percent of STZ- Complete Blood Cell Count injected dams (27/30) had mean glucose levels greater than 11 mmol/l Blood was sampled from 6- and 11-mo-old OCM and ODM in ethyl- (200 mg/dl). Sixty percent of dams (18/30) received at least one injec- enediaminetetraacetic acid-coated tubes (Sarstedt, Newton, NC), and tion of insulin glargine (1U, Lantus, Sanofi-Aventis, Bridgewater, NJ) or leukocytes quantitated using a Hemavet (Drew Scientific, Waterbury, Humulin (1U, Eli Lilly, Indianapolis, IN) subcutaneously when glucose CT). levels were above 28 mmol/l (500 mg/dl). Pups were born spontaneously, and crossfostered to noninjected mothers in litters of 10 pups (5 male Bone Marrow Progenitor Colony Assays and 5 female) at postnatal day 1. At 3 wk of age, animals were weaned To obtain bone marrow, femurs from 6- and 11-mo-old rats were and fed either a normal diet (14% kcal from fat, Teklad 7001, Harlan flushed with Iscove’s modified Dulbecco’s medium (Life Technologies, Laboratories, Madison, WI) or a HFD (60% kcal from fat, TD.06414, Grand Island, NY) containing 20% fetal bovine serum (Atlanta Teklad, Harlan Laboratories) ad libitum. Animals were weighed weekly Biologicals, Flowery Branch, GA). Cells were washed and counted and were group housed until they reached weights of 500 g. One cohort using a Coulter counter (Beckman Coulter, Brea, CA) to deter- of animals was euthanized at 6 mo of age, at which time complete blood mine total femur cellularity. Cells (50,000 per 35-mm dish) were cell counts were performed, and plasma and bone marrow were col- plated in triplicate in 40% methylcellulose media (M3134, Stem Cell lected (Figure 6). A separate cohort of rat offspring were allowed to con- Technologies, Vancouver, Canada), 45% fetal bovine serum, 0.69 sume normal diet until the age of 9 mo, when they were started on the nmol/l granulocyte macrophage colony-stimulating factor (10 ng/ HFD for 2 mo (Figure 6). Aged animals were euthanized at 11 mo of age ml mouse granulocyte-macrophage colony-stimulating factor, at which time complete blood cell counts were performed, and plasma Peprotech, Rocky Hill, NJ) and 2.7 nmol/l stem cell factor (50 ng/ml and bone marrow were collected. stem cell factor; Peprotech), 2 U/ml erythropoietin (Epogen, Amgen, Thousand Oaks, CA), 78 μmol/l β-mercaptoethanol (Fisher Scientific, Quantitation of Pancreatic β Cell Area Pittsburgh, PA), and penicillin-streptomycin (Mediatech, Manassas, Pancreata were isolated from control and STZ-treated dams after VA). After 7 d of incubation at 37 °C, total colonies were counted to delivery of pups, and from neonatal male and female rats on post- determine the frequency of progenitors. Total progenitors per femur natal day 1. Specimens were fixed in formalin overnight, embedded were calculated as the product of the progenitor frequency and the in paraffin, and sectioned by the Histology Core of the Department total cells per femur as previously described (49). of Anatomy and Cell Biology at the Indiana University School of Medicine. Staining for insulin was performed (sc-9168, 1:500, Santa Data Analysis Cruz Biotechnology, Dallas, TX), with detection using Impress Statistical analyses and graphs were created using GraphPad Prism anti-rabbit horseradish peroxidase and Novared substrate (Vector 6 software (GraphPad Software, San Diego, CA). Data reported in Laboratories, Burlingame, CA). A Zeiss Axio Observer Z1 inverted the text are mean ± SD. Statistical analyses were performed within microscope (Zeiss, Thornwood, NY) equipped with an Orca ER CCD a specific sex group as indicated in the figure legends and tables, not camera (Hamamatsu Photonics, Hamamatsu City, Japan) was used between males and females. Statistical significance was determined to acquire digital images of the entire stained longitudinal pancreatic by unpaired t-test, or by one-way, two-way, or repeated measures section. The β cell area of at least five sections, each separated by at ANOVA with Tukey’s or Sidak’s multiple comparison tests as speci- least 75 µm, from at least three animals in each group was calculated fied in the figure legends and tables. A χ2 test was used to analyze using Axio-Vision Software (Zeiss) as previously described (48). frequency of maternal deaths during labor.

Glucose Tolerance Test ACKNOWLEDGMENTS After a 16-h fast, animals were injected with glucose (2 g/kg i.p.) at the The authors thank Shehnaz Khan for technical support on this project. We ages of 4, 8, and 13 wk. Glucose levels were monitored from the tail also thank Elizabeth Rybak for providing excellent administrative support. vein using an Accu-Check Aviva glucose monitor (Roche). Plasma samples were obtained for insulin measurements in a separate cohort STATEMENT OF FINANCIAL SUPPORT of animals at earlier time points. Plasma was isolated and stored at This study was supported by The Riley Children’s Foundation, Indianapolis, −80 °C until used for radioimmunoassays. Since 11-mo-old rats on a USA (L.S.H.); U.S. National Institutes of Health (Bethesda, MD), R01 DK076912 HFD had a qualitative increase in abdominal adiposity, intravenous (D.C.T.), K08 DK080225 (C.E.M.), R03 DK089147 (C.E.M.), and R01 DK093954 glucose tolerance tests were conducted under anesthesia to enhance (C.E.M.); and VA Merit Award 1I01BX001733 (Washington, DC) (C.E.M.). reliability of glucose administration and decrease procedural stress for these animals. In the 11-mo-old animals, the rats were fasted Disclosure: The authors declared no conflict of interest. and anesthetized with isoflurane before tail vein injection of glucose

(0.5g/kg). Glucose levels were monitored from the tail vein using an References Accu-Check Aviva glucose monitor. 1. American Diabetes Association. Standards of medical care in diabe- Insulin Tolerance Test tes--2014. Diabetes Care 2014;37:Suppl 1:S14–80. After a 2-h fast, 8-wk-old animals were injected with insulin (0.75 U/ 2. Metzger BE, Buchanan TA, Coustan DR, et al. Summary and recommen- kg i.p., Humulin, Eli Lilly) and tail vein blood glucose levels were dations of the Fifth International Workshop-Conference on Gestational measured using an Accu-Check Aviva glucose monitor (Roche). Diabetes Mellitus. Diabetes Care 2007;30:Suppl 2:S251–60.

3. Dabelea D, Snell-Bergeon JK, Hartsfield CL, Bischoff KJ, Hamman RF, 27. DiMascio L, Voermans C, Uqoezwa M, et al. Identification of adiponec- McDuffie RS; Kaiser Permanente of Colorado GDM Screening Program. tin as a novel hemopoietic stem cell growth factor. J Immunol 2007;178: Increasing prevalence of gestational diabetes mellitus (GDM) over time 3511–20. and by birth cohort: Kaiser Permanente of Colorado GDM Screening Pro- 28. Claycombe K, King LE, Fraker PJ. A role for leptin in sustaining lympho- gram. Diabetes Care 2005;28:579–84. poiesis and myelopoiesis. Proc Natl Acad Sci USA 2008;105:2017–21. 4. Weintrob N, Karp M, Hod M. Short- and long-range complications in off- 29. Bennett BD, Solar GP, Yuan JQ, Mathias J, Thomas GR, Matthews W. spring of diabetic mothers. J Diabetes Complications 1996;10:294–301. A role for leptin and its cognate receptor in hematopoiesis. Curr Biol 5. Boney CM, Verma A, Tucker R, Vohr BR. Metabolic syndrome in child- 1996;6:1170–80. hood: association with birth weight, maternal obesity, and gestational dia- 30. Fantuzzi G, Faggioni R. Leptin in the regulation of immunity, inflamma- betes mellitus. Pediatrics 2005;115:e290–6. tion, and hematopoiesis. J Leukoc Biol 2000;68:437–46. 6. Bunt JC, Tataranni PA, Salbe AD. Intrauterine exposure to diabetes is a 31. Gainsford T, Willson TA, Metcalf D, et al. Leptin can induce proliferation, determinant of hemoglobin A(1)c and systolic blood pressure in pima differentiation, and functional activation of hemopoietic cells. Proc Natl Indian children. J Clin Endocrinol Metab 2005;90:3225–9. Acad Sci USA 1996;93:14564–8. 7. Cho NH, Silverman BL, Rizzo TA, Metzger BE. Correlations between the 32. Bush NC, Chandler-Laney PC, Rouse DJ, Granger WM, Oster RA, intrauterine metabolic environment and blood pressure in adolescent off- Gower BA. Higher maternal gestational glucose concentration is associ- spring of diabetic mothers. J Pediatr 2000;136:587–92. ated with lower offspring insulin sensitivity and altered beta-cell function. 8. Wright CS, Rifas-Shiman SL, Rich-Edwards JW, Taveras EM, Gillman MW, J Clin Endocrinol Metab 2011;96:E803–9. Oken E. Intrauterine exposure to gestational diabetes, child adiposity, and 33. Chandler-Laney PC, Bush NC, Granger WM, Rouse DJ, Mancuso MS, blood pressure. Am J Hypertens 2009;22:215–20. Gower BA. Overweight status and intrauterine exposure to gestational 9. Blondeau B, Joly B, Perret C, et al. Exposure in utero to maternal diabetes diabetes are associated with children’s metabolic health. Pediatr Obes leads to glucose intolerance and high blood pressure with no major effects 2012;7:44–52. on lipid metabolism. Diabetes Metab 2011;37:245–51. 34. Mehran AE, Templeman NM, Brigidi GS, et al. Hyperinsulinemia drives 10. Ding GL, Wang FF, Shu J, et al. Transgenerational glucose intolerance with diet-induced obesity independently of brain insulin production. Cell Igf2/H19 epigenetic alterations in mouse islet induced by intrauterine Metab 2012;16:723–37. hyperglycemia. Diabetes 2012;61:1133–42. 35. Gray SL, Donald C, Jetha A, Covey SD, Kieffer TJ. Hyperinsulinemia pre- 11. Bond H, Hamilton K, Balment RJ, et al. Diabetes in rat pregnancy alters cedes insulin resistance in mice lacking pancreatic beta-cell leptin signal- renal calcium and magnesium reabsorption and bone formation in adult ing. Endocrinology 2010;151:4178–86. offspring. Diabetologia 2005;48:1393–400. 36. Shanik MH, Xu Y, Skrha J, Dankner R, Zick Y, Roth J. Insulin resistance 12. Zhao J, Weiler HA. Long-term effects of gestational diabetes on offspring and hyperinsulinemia: is hyperinsulinemia the cart or the horse? Diabetes health are more pronounced in skeletal growth than body composition and Care 2008;31:Suppl 2:S262–8. glucose tolerance. Br J Nutr 2010;104:1641–9. 37. Fonseca SG, Urano F, Burcin M, Gromada J. Stress hypERactivation in the 13. Nehiri T, Duong Van Huyen JP, Viltard M, et al. Exposure to maternal β-cell. Islets 2010;2:1–9. diabetes induces salt-sensitive hypertension and impairs renal function in 38. Eizirik DL, Cardozo AK, Cnop M. The role for endoplasmic reticulum adult rat offspring. Diabetes 2008;57:2167–75. stress in diabetes mellitus. Endocr Rev 2008;29:42–61. 14. Holemans K, Gerber RT, Meurrens K, De Clerck F, Poston L, Van Ass- 39. Park JH, Stoffers DA, Nicholls RD, Simmons RA. Development of type che FA. Streptozotocin diabetes in the pregnant rat induces cardiovascular 2 diabetes following intrauterine growth retardation in rats is associated dysfunction in adult offspring. Diabetologia 1999;42:81–9. with progressive epigenetic silencing of Pdx1. J Clin Invest 2008;118: 15. Gauguier D, Bihoreau MT, Picon L, Ktorza A. Insulin secretion in adult 2316–24. rats after intrauterine exposure to mild hyperglycemia during late gesta- 40. Schnedl WJ, Ferber S, Johnson JH, Newgard CB. STZ transport and tion. Diabetes 1991;40:Suppl 2:109–14. cytotoxicity. Specific enhancement in GLUT2-expressing cells. Diabetes 16. Agoudemos M, Reinking BE, Koppenhafer SL, Segar JL, Scholz TD. Pro- 1994;43:1326–33. gramming of adult cardiovascular disease following exposure to late-gesta- 41. Hathout EH, Kumagai AK, Sangkharat A, Geffner ME, Mullen Y. Absence tion hyperglycemia. Neonatology 2011;100:198–205. of GLUT2 protein in near-term fetal rat pancreatic islets. Pancreas 17. Katkhuda R, Peterson ES, Roghair RD, Norris AW, Scholz TD, Segar JL. 1997;14:318–21. Sex-specific programming of hypertension in offspring of late-gestation 42. Hughes SJ. The role of reduced glucose transporter content and glucose diabetic rats. Pediatr Res 2012;72:352–61. metabolism in the immature secretory responses of fetal rat pancreatic 18. Segar EM, Norris AW, Yao JR, et al. Programming of growth, insulin resis- islets. Diabetologia 1994;37:134–40. tance and vascular dysfunction in offspring of late gestation diabetic rats. 43. Mansuy-Aubert V, Zhou QL, Xie X, et al. Imbalance between neutro- Clin Sci (Lond) 2009;117:129–38. phil elastase and its inhibitor α1-antitrypsin in obesity alters insulin 19. Zinkhan EK, Fu Q, Wang Y, et al. Maternal hyperglycemia disrupts sensitivity, inflammation, and energy expenditure. Cell Metab 2013;17: histone 3 lysine 36 trimethylation of the IGF-1 gene. J Nutr Metab 534–48. 2012;2012:930364. 44. Talukdar S, Oh da Y, Bandyopadhyay G, et al. Neutrophils mediate insulin 20. Centers for Disease Control and Prevention. National Diabetes Fact Sheet, resistance in mice fed a high-fat diet through secreted elastase. Nat Med 2011. http://www.cdc.gov/diabetes/pubs/estimates11.html. 2012;18:1407–12. 21. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of Obesity in the United 45. Mócsai A. Diverse novel functions of neutrophils in immunity, inflamma- States, 2009–2010. http://www.cdc.gov/nchs/data/databriefs/db82.pdf. tion, and beyond. J Exp Med 2013;210:1283–99. 22. Elgazar-Carmon V, Rudich A, Hadad N, Levy R. Neutrophils transiently 46. Weisberg SP, McCann D, Desai M, Rosenbaum M, Leibel RL, Fer- infiltrate intra-abdominal fat early in the course of high-fat feeding. J Lipid rante AW Jr. Obesity is associated with macrophage accumulation in adi- Res 2008;49:1894–903. pose tissue. J Clin Invest 2003;112:1796–808. 23. Xu H, Barnes GT, Yang Q, et al. Chronic inflammation in fat plays a crucial 47. Lu Z, Wang Z, Wang X, et al. Protection from high-fat-diet-induced role in the development of obesity-related insulin resistance. J Clin Invest impaired glucose tolerance in female Sprague-Dawley rats. Gynecol Endo- 2003;112:1821–30. crinol 2009;25:464–71. 24. Lumeng CN, Bodzin JL, Saltiel AR. Obesity induces a phenotypic switch in 48. Sims EK, Hatanaka M, Morris DL, et al. Divergent compensatory responses adipose tissue macrophage polarization. J Clin Invest 2007;117:175–84. to high-fat diet between C57BL6/J and C57BLKS/J inbred mouse strains. 25. Szkudelski T. The mechanism of alloxan and streptozotocin action in B Am J Physiol Endocrinol Metab 2013;305:E1495–511. cells of the rat pancreas. Physiol Res 2001;50:537–46. 49. Haneline LS, Broxmeyer HE, Cooper S, et al. Multiple inhibitory cytokines 26. Borregaard N. Neutrophils, from marrow to microbes. Immunity induce deregulated progenitor growth and apoptosis in hematopoietic cells 2010;33:657–70. from Fac-/- mice. Blood 1998;91:4092–8.